Development of low complexity evolutionary computing based nonlinear active noise control systems

Abstract

A nonlinear active noise control (ANC) system based on a couple of low complexity nonlinear networks are developed in this paper. These are the evolutionary computing based feed forward nonlinear network (FFNN) and the evolutionary computing based feed forward recursive nonlinear network (FFRNN). The new method does not require the identification of the secondary path, which not only improves the stability of the ANC system but also reduces the computational complexity. The design of the proposed ANC systems is viewed as a single objective optimization problem in which the weights of the ANC system are updated using particle swarm optimization (PSO) based evolutionary algorithm.